Drug Class Review Pharmacologic Treatments for Attention Deficit Hyperactivity Disorder Final Report Update 3 September 2009 Update 2: November 2007 Update 1: May 2006 Original Report: September 2005 The literature on this topic is scanned periodically. This report reviews information about the comparative effectiveness and safety of drugs within a pharmaceutical class. The report is neither a usage guideline nor an endorsement or recommendation of any drug, use, or approach. Oregon Health & Science University does not endorse any guideline or recommendation developed by users of this report. Marian S. McDonagh, PharmD Vivian Christensen, PhD Kim Peterson, MS Sujata Thakurta, MPA:HA Marian McDonagh, PharmD, Principal Investigator Oregon Evidence-based Practice Center Mark Helfand, MD, MPH, Director Oregon Health & Science University Copyright 2009 by Oregon Health & Science University Portland, Oregon 97239. All rights reserved.
INTRODUCTION According to the most recent National Institutes of Health Consensus Statement (1998), attention deficit hyperactivity disorder is the most commonly diagnosed childhood behavioral disorder. Classification of hyperactivity and defects in attention emerged in the 1960 s as Minimal Brain Dysfunction and Hyperkinetic Syndrome, and has continued to evolve over time. A number of community-based studies have reported attention deficit hyperactivity disorder (ADHD) prevalence rates that range from 1.7% to 16%. This is broader than the range of 3% to 5% that was estimated by the expert panelists that participated in the National Institutes of Health Consensus Development Conference on Diagnosis and Treatment of Attention Deficit Hyperactivity Disorder in 1998. The estimated prevalence cited in the most recent (1997) version of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) is 3% to 7%. Differences in prevalence estimates may be due to variation in methods of ascertainment and diagnostic criteria. While no independent diagnostic test exists for ADHD, the DSM-IV provides standardized criteria that can be used as a foundation for clinical diagnosis. According to the DSM-IV, essential features of ADHD include persistent levels of inattention, impulsivity, and/or hyperactivity that exceed usual developmental patterns. In order to qualify for a DSM-IV diagnosis of ADHD, symptoms must date back to before age 7, persist for at least 6 months, and cause impairment that interferes with functional capacity in at least 2 performance settings (social, academic, or employment). DSM-IV specifies 3 distinct subtypes of ADHD that are characterized by predominantly inattentive, hyperactive-impulsive, or mixed symptoms. Scope and Key Questions The purpose of this review is to compare the benefits and harms of different pharmacologic treatments for ADHD. Included drugs are described in Table 1. Table 1. ADHD drugs and indication (immediate-release and extended-release formulations) Referred to in this Active ingredient(s) summary as Trade name a Forms Amphetamine mixture Adderall a,b MAS IR Oral tablet (amphetamine aspartate; amphetamine sulfate; dextroamphetamine saccharate; MAS XR Adderall XR Extended-release dextroamphetamine sulfate) Atomoxetine HCl Atomoxetine Strattera Oral capsule d-mph IR Focalin a,b Oral tablet Dexmethylphenidate b Extended-release hydrochloride d-mph ER Focalin XR Dexedrine a Oral tablet DEX IR Dextrostat a,d Oral tablet Dextroamphetamine sulfate Liquadd Oral solution Dexedrine Sustained-release DEX SR Spansule Lisdexamfetamine dimesylate Lisdexamfetamine Vyvanse d Oral capsule Methamphetamine hydrochloride Methamphetamine Desoxyn b Oral tablet Attention deficit hyperactivity disorder Page 2 of 9
Active ingredient(s) Methylphenidate hydrochloride Modafinil Referred to in this summary as Trade name a Forms MPH OROS Concerta Extended-release oral tablet b Transdermal MPH transdermal Daytrana patch b Extended-release MPH CD Metadate CD MPH ER Metadate ER b Extended-release oral tablet Medikinet c Extended-release MPH chewable MPH solution a Or generic equivalent. b Not available in Canada. c Not available in the United States. d Approved in Canada but not commercially available. Methylin b oral tablet Oral chewable tablet Oral solution MPH IR Ritalin a Oral tablet b Extended-release MPH LA Ritalin LA c Extended-release Multi-layer MPH Biphentin MPH SR Ritalin SR Extended-release oral tablet Provigil Oral tablet Modafinil Alertec c Oral tablet The participating organizations of the approved the following key questions to guide this review: 1. Evidence on Effectiveness and Efficacy a. What is the comparative or noncomparative evidence that pharmacologic treatments for attention deficit disorders improve effectiveness outcomes? b. What is the comparative efficacy of different pharmacologic treatments for attention deficit disorders? 2. Tolerability, Serious Adverse Events, Misuse and Diversion a. What is the evidence of comparative tolerability of different pharmacologic treatments for attention deficit disorders? b. What is the evidence of serious adverse events or long-term adverse events associated with use of pharmacologic treatments for attention deficit disorders? c. What is the comparative or noncomparative evidence that pharmacologic treatments for attention deficit disorders impact the risk of misuse or illicit diversion in patients with no history of misuse or diversion? i. Stimulants compared with nonstimulants ii. Immediate release compared with intermediate compared with long-acting formulations iii. Any included pharmacologic treatment Attention deficit hyperactivity disorder Page 3 of 9
3. Evidence in Subgroups of Patients a. What is the evidence of benefits and harms of pharmacologic treatments for attention deficit disorders in subgroups of patients based on demographics (age, racial groups, gender), other medications or therapy, or comorbidities (e.g. tics, anxiety, substance use disorders, disruptive behavior disorders)? b. What is the comparative or noncomparative evidence of misuse or illicit diversion of pharmacologic treatments for attention deficit disorders in patients with current or past substance use disorder comorbidities? i. Stimulants compared with nonstimulants ii. Immediate release compared with intermediate compared with long-acting formulations iii. Any included pharmacologic treatment METHODS Literature Search To identify relevant citations, we searched the Cochrane Central Register of Controlled Trials (1st Quarter 2009), Cochrane Database of Systematic Reviews (1st Quarter 2009), MEDLINE (1996 to April Week 4 2009), and PsycINFO (1806 to April Week 4 2009) using terms for included drugs, indications, and study designs. We have attempted to identify additional studies through searches of reference lists of included studies and reviews, the US Food and Drug Administration web site, as well as searching dossiers submitted by pharmaceutical companies for the current review. Validity Assessment We assessed the internal validity (quality) of trials based on the United States Preventive Services Task Force and the National Health Service Centre for Reviews and Dissemination (U.K.) criteria. We rated the internal validity of each trial based on the methods used for randomization, allocation concealment, and blinding; the similarity of compared groups at baseline; maintenance of comparable groups; adequate reporting of dropouts, attrition, crossover, adherence, and contamination; loss to follow-up; and the use of intention-to-treat analysis. Trials that had a fatal flaw in one or more category were rated poor-quality ; trials that met all criteria were rated good-quality ; the remainder were rated fair-quality. A fatal flaw occurs when there is evidence of bias or confounding in the trial, for example when randomization and concealment of allocation of random order are not reported and baseline characteristics differ significantly between the groups. In this case, randomization has apparently failed and for one reason or another bias has been introduced. Attention deficit hyperactivity disorder Page 4 of 9
RESULTS Overview Overall, we included 369 studies, 71 of these were added in Update 3. Of these, 69 were direct comparisons of one drug versus another in a randomized, controlled trial. Dossiers were submitted by Eli Lilly (atomoxetine HCl), Shire US (lisdexamfetamine dimesylate and transdermal methylphenidate), and McNeil (methylphenidate OROS) for the most recent update of this report. There were no trials of comparative effectiveness of these drugs for treatment of ADHD and good-quality evidence on the use of drugs to affect outcomes relating to global academic performance, consequences of risky behaviors, social achievements, etc. was lacking. The evidence for comparative efficacy and adverse events of drugs for treating ADHD was severely limited by small sample sizes, very short durations, and the lack of studies measuring functional or long-term outcomes. Methods of measuring symptom control vary significantly across studies. The crossover design was frequently used, with few analyzing the effect of order of administration of drugs. Those that did found a significant effect. No head-to-head efficacy trial was good quality. The small numbers of patients in these trials limited the ability to show a difference between drugs if one exists. Limitations to the generalizability of these trials included limited characterization of ADHD symptomatology across studies due to use of varied or indeterminate diagnostic processes and underrepresentation of minorities and the most seriously ill patients. The small sample sizes of these trials did not allow for statistical analyses of potential effects of these factors. Overall, the rate of response to stimulants appeared to be in the range of 60% to 80%, however the definitions of response rate varied and may not be comparable. Depending on the definition used, there is lack of clarity on the relationship of response rate to clinical significance. Response rates of nonstimulants varied, but the range in placebo-controlled trials was similar to that found with stimulants. Significant variation in the method of assessment and definition of response was most likely the reason for the wide variation. SUMMARY Results by key question are summarized in Table 2, below. Table 2. Summary of the evidence Key Question 1. Benefits General Effectiven No trials found: Poor ess Young children Efficacy Overall: Poor MPH IR Children No conclusions about comparative effectiveness of different drugs for ADHD can be made. The evidence on efficacy of MPH IR in the short term is mixed. Attention deficit hyperactivity disorder Page 5 of 9
Efficacy Stimulants IR vs. SR formulatio ns SR vs. SR formulatio ns IR vs. IR Overall: (individual ratings below) MPH IR vs. MPH SR: MPH SR vs. MPH SR formulations: Poor DEX IR vs. MPH IR: Good MAS IR vs. MPH IR: Studies of MPH IR vs. extended release formulations in children generally were unable to identify significant differences in symptom improvement. Studies of MPH IR and MPH OROS are conflicting; a difference was not found in double-blind studies while open-label studies indicate greater improvement with MPH OROS on some measures. Limited evidence from 2 small crossover studies suggests that MPH LA was superior to MPH OROS on some, but not all efficacy outcomes. Limited evidence suggests that MPH CD was superior to MPH OROS on outcomes in the morning; they had similar effects in the afternoon; and MPH OROS was superior in the evening. d-mph ER was superior to MPH OROS at 2 to 6 hours post-dose, and MPH OROS was superior at 10 to 12 hours in 1 trial. The body of evidence clearly indicates no difference in efficacy between DEX and MPH IR. MAS IR was superior to MPH IR on a few efficacy outcome measures in 2 trials, but clear evidence of superiority is lacking. Transdermal MPH Lisdexamfet amine Atomoxetine Adolescents Efficacy Adults Efficacy Direct comparis ons DEX IR vs. DEX ER vs. MAS: Poor Modafinil vs. MPH IR: Dexmethylphenidate: NA Transdermal MPH vs. MPH OROS Poor Atomoxetine vs. MPH IR Atomoxetine vs. MAS XR Atomoxetine vs. MPH OROS Poor MPH OROS vs. MAS IR MPH IR vs. MPH OROS Placebo-controlled studies of MPH IR DEX IR vs. modafinil Evidence on the comparison of DEX IR vs. SR vs. MAS may suggest that measures made in the morning show DEX IR superior to DEX SR, and afternoon measures show DEX SR superior to MAS. Based on 1 trial, modafinil was similar to MPH IR in efficacy Only placebo-controlled evidence was found. Based on 1 trial, MTS and MPH OROS had similar efficacy Lisdexamfetamine was comparable to MAS XR on average SKAMP-DS scores and superior to placebo on same, as well as on ADHD rating scale IV mean changes. Limited evidence suggests a lack of a difference in efficacy compared to MPH IR. Limited evidence suggests that MAS XR is superior to atomoxetine on most efficacy measures. MPH OROS was superior to atomoxetine in response rates Effectiveness outcomes: NR Short-term improvements in core ADHD symptoms: No differences. Other: MPH OROS > MAS IR on overall simulator driving performance. Functional capacity: NR Short-term improvements of core ADHD symptoms: NR. Driving performance: MPH OROS > MPH IR in evening and at night. Functional capacity: NR Short-term improvements of core ADHD symptoms: MPH IR generally efficacious. Limited evidence suggests a lack of a difference in efficacy between DEX IR and modafinil. Attention deficit hyperactivity disorder Page 6 of 9
Indirect comparis ons Key Question 2. Safety Atomoxetine, DEX IR, d- MPH XR, lisdexamfetamine, MPH ER, MPH IR, MPH SR, MPH OROS, MAS IR, MAS XR: d-mph IR, MPH transdermal patch, Metadate CD, Ritalin LA, and Biphentin ; Poor All were found to be effective short-term treatments for reducing ADHD symptoms in placebo-controlled trials. Pooled analyses suggest a relative benefit of clinical response for shorter acting stimulants at 3.26 times greater than for patients taking longeracting stimulants (95% CI, 2.03 to 5.22). Atomoxetine: Not consistently significantly superior to placebo in improving quality of life and driving performance outcomes MPH IR: Consistently superior to placebo in improving driving performance outcomes. MAS XR: Superior to placebo in improving overall simulated driving performance in 1 trial No evidence. 2b. Short-term trial evidence Young children 1 placebo-controlled trial of MPH: Poor Indirect comparisons cannot be made; MPH associated with higher rates of adverse events than placebo. Children Poor Very few studies reported methods for assessing adverse events a priori. MPH IR vs. MPH SR There is no evidence of a difference in adverse events between IR and SR formulations. MPH SR vs. MPH SR No differences in adverse events were found. formulations DEX vs. MPH IR Limited evidence from short-term trials suggests that weight loss is greater with DEX than MPH IR. MAS vs. MPH IR Very limited evidence suggests that twice daily dosing of MAS led to higher rates of loss of appetite and sleep trouble. DEX IR vs. DEX ER vs. MAS Transient weight loss was greater with MAS and DEX SR than with DEX IR. Comparisons to atomoxetine Rates of vomiting ranged from 12% to 13% for atomoxetine, which was approximately 3 times greater than rates for MPH IR or MAS XR. Rates of somnolence ranged from 6% to 26% for atomoxetine, which was 3 to 4 times greater than rates for longer-acting stimulants (MPH OROS and MPH XR) and over 7 times greater than rates in trials of MPH IR. Nausea and anorexia were greater with atomoxetine compared to MPH IR in 1 trial. MPH OROS and MAS XR caused higher rates of insomnia (7% atomoxetine, 13% MPH OROS, 28% MAS XR) than atomoxetine in 2 trials. Lisdexamfetamine No differences in adverse event rates between lisdexamfetamine vs. MAS XR. Teens Poor Very few studies reported methods for assessing adverse events a priori. Placebo-controlled studies of MPH IR No indirect comparisons possible. Placebo-controlled trials only involved assessment of MPH IR. Adults Poor Very few studies reported methods for assessing adverse events a priori. Rates of appetite disturbance and sleep disturbance were generally greater for atomoxetine, DEX IR, d-mph-er, lisdexamfetamine, MPH ER, MPH IR, MPH SR, MPH OROS, MAS IR, and MAS XR Our adjusted indirect meta-analysis found that shorter-acting stimulants, longer-acting stimulants, and atomoxetine groups had significantly higher risk of appetite loss and sleep disturbance relative to placebo, but indirect comparisons suggest no significant difference between drug types. Attention deficit hyperactivity disorder Page 7 of 9
Adderall and MPH IR DEX IR and MPH SR Atomoxetine Indirect comparisons from placebo-controlled trials suggest both are associated with higher rates of insomnia, appetite loss and withdrawal due to adverse events than placebo. Indirect comparisons cannot be made. Very limited indirect comparative evidence across few placebo-controlled trials suggests that atomoxetine is associated with rates of insomnia, appetite loss and withdrawals due to adverse events similar to stimulants. 2b. Long-term safety: Observational studies Mixed populations, primarily children Sudden cardiac death Increased risk associated with current stimulant use (odds ratio 7.4; 95% CI, 1.4 to 74.9) based on case control study. Smaller study found no association. Recall bias may be an issue. Cardiac events Emergency room and physician office visits for cardiac causes significantly more frequent among those taking stimulants compared with those not (hazard ratio, 1.20; 95% CI, 1.04 to 1.38 compared with hazard ratio, 1.21; 95% CI, 1.06 to 1.30). Suicidal behavior Increased risk with atomoxetine compared to placebo (risk difference, 0.52; 95% CI, 0.12 to 0.91) based on meta-analysis. Time to onset of behavior 9 to 32 days. Overall rate of suicidal behavior and ideation was 0.44% in this study compared to 1.7% in another meta-analysis of longer-term duration. Height DEX vs. MPH IR: Mixed findings. DEX=MPH in 6-year height increases in 1 study; DEX>MPH in 2-year height decreases in the other. MPH IR vs. unmedicated controls: No significant differences in 2 studies. MPH IR in uncontrolled studies: Inconsistent effects across 4 studies. Atomoxetine: Uncontrolled studies suggest that height changes are similar to those reported with MPH IR, and are also transient. Weight DEX vs. MPH: Three studies consistently suggest that DEX>MPH in weight gain suppression in the first 1-2 years. The longest-term (5 years) of these studies also reported that DEX=MPH in exceeding weight gain expectations at final follow-up. These findings are weakened by methodological flaws, however. MPH IR in other comparative (imipramine and unmedicated hyperactives or healthy controls) and noncomparative studies: Evidence does not support an indisputable relationship between MPH and weight gain suppression. MPH OROS and tomoxetine (atomoxetine): Evidence from noncomparative studies (1 each) doesn t suggest weight gain suppression effects. Atomoxetine: Uncontrolled studies suggest that weight changes are similar to those reported with MPH IR, and are also transient. Tics, seizures, No comparative evidence. cardiovascular adverse events, injuries, and suicidal behavior 2c. Abuse/diversion Teens and young adults Poor Stimulant use during childhood not associated with alcohol abuse later. May be protective against or delay nicotine dependence, but comorbid conduct disorder may be a significant confounder. Stimulant use may protect against later substance abuse, but again comorbid conduct disorder may be a confounder. Evidence on misuse and diversion reports wide ranges of prevalence with no comparative data. Attention deficit hyperactivity disorder Page 8 of 9
Key Question 3. Subgroups Children ADHD subtypes or severity Race/ethnicity Gender Tic disorders Oppositional defiant disorder Bipolar disorder Abbreviations: ADHD, attention deficit hyperactivity disorder. Atomoxetine, MPH IR, MPH OROS all have superior efficacy relative to placebo in children with ADHD, regardless of diagnostic subtype but response may be better in those with combined or inattentive subtype. Most trials conducted in primarily White populations. Ethnicity/race only reported in one half of studies. No analyses based on race. Very limited evidence suggests MPH IR in African American boys results in response rates similar to other populations studied. Evidence from subgroup analysis of a placebo-controlled trial suggested that effects of lisdexamfetamine may be less robust in non-caucasian children. Subgroup analyses based on gender were limited. Evidence from subgroup analysis of a placebo-controlled trial suggested that lisdexamfetamine may be less efficacious in girls. Exploratory analysis indicates atomoxetine may have better response on emotional regulation items in women than men. No consistent evidence that atomoxetine, DEX IR or MPH IR increased tic severity or frequency compared to placebo. All of these studies of MPH IR showed a benefit of MPH IR on ADHD outcome measures compared to placebo. Very limited evidence suggests that atomoxetine is beneficial on most ADHD outcomes compared to placebo. Very limited evidence suggests that MAS IR or MPH IR have benefit on most ADHD outcomes compared to placebo. Attention deficit hyperactivity disorder Page 9 of 9